I'm trying to sort the map I'm going to use.

The original has this as the plan



But I don't think Europe or the UK is detailed enough so I've knocked up this which is a bit more detailed. Hopefully it's not too difficult to actually drill. It'll be a lot more holes.



Though this might be better as a compromise


 

That looks like far too many holes to drill (UYM,J)

I did consider counting them on the way up yo Stoke, going to see Top Gear Live in Liverpool tomorrow, but I decided against it. Been dragged out for drinks with my drunk sister and friends. And they've all desserted me for the toilet.

Started soldering the main board.

So far I think I've done a reasonable job, but it's only the easy bits. Felt like I was cheating using the long pins of the transistors to make things easier!

So far I've learnt that dip sockets are quite easy to solder, female pin headers are an absolute pain.

I've finished the board. But it's not working properly :C

The second and third shift registers aren't working. They seem to cause the LEDs to be full brightness . I've checked the wiring but I think I need to check again with a fresh mind. Though I did read something about communication issues across shift registers with poor wiring.

Any luck with it?

Not had chance (been bothered) to have a proper sit down with it. Bought a cellulose decorating sponge the other day, so I finally have sponge!

Removing all of that multi-coloured capellini laying all over the board should help a lot.

It needs to taste good though!

Arduino... The taste you have been looking for.

It seems the step up reg might not be the best way after all. Either that or the Pro Mini is consuming more power than I realise. Got about a week from the last set of batteries.

New plan is to drop the clock speed so it can run at 1.8V and switch the step up regulator on when the temperature is read. The sensor needs 3V.

I've also ordered a couple of Atmega328p chips to make a barebones board.

And my Nanos have /finally/ arrived. Got one of them controlling my bedroom LEDs now. Borrowed the mini that I'd set up to do it before.

Version 0.2 done. Hoping to get far more battery life out of this one. I have a little test module running just sending it's battery state every 2 minutes and after 3 days it's only dropped ~0.004V

Moved a couple of cables around on the board, couple of optimisations to the code so it only sends if the temp has changed (or every 10 minutes). Switching the power to the radio and temp sensor off completely when sleeping. Only problem is the DS18B20 requires a little time to start up and read the temperature so it's running for almost 1 second when it wakes up.

And it /should/ work all the way down to 1.8V as it's now running at 4MHz. I'm surprised the radio works as I read it can be fussy with timings.

I just built this. Going to use it for testing various methods of powering my devices. Quick swap of a couple of leads and I can change between battery or step up or switched from an output.

Well that's no good. The temp module's battery is dropping quickly again. I estimate maybe 6 days from it. Thought it would do better. Think I might attach an LED to it so I can see how long it's waking for. Unfortunately pin 13 doesn't work because the SPI uses it.

I bought an ATMega328p chip a while back to try and make a barebones board but couldn't get it working. Also, the price of the parts was far more than buying a Nano or Pro Mini so it seemed a little pointless.

I have seen some people running strip down boards off button cells for a very long time, usually only transmitting something occasionally though.

I might try a 433 MHz transmitter on a temp sensor setup to see how well that works.

http://jeelabs.org/tag/lowpower/

I think it /should/ be easy to get working at 8MHz. Shouldn't need many components, maybe just a cap. Maybe not so easy to sort out the boot loader. It worked out slightly cheaper to buy the chip than a Pro Mini, about 20p I think. Amazing they can sell a fully assembled board for only slightly more. Maybe the DIPs are more expensive.

I have another Pro Mini just sending it's battery voltage. That should last months. So something up with the other one. Either somethring to do with the temp sensor or the Mini.

Is there a less power hungry way to read the temperature? What about a pressure sensor in a sealed tube and use P1/T1 = P2/T2 to sovle from a reference temp?

I imagine it's the wireless transmitting of the data which uses more power out of everything.

(I think) The problem is the sensor itself. It needs at least 3V so I'm using a step up to increase the battery voltage. I'm switching the power off to it when the Arduino goes sleep. When it powers on again I get a temp of 85C unless I delay the reading by about half a second, so it's staying awake too long each cycle. The actual sensor uses very little.

I wonder if leaving it on would be better. Will leaving the step up on waste much? My other option is to switch the sensor on, go to sleep for 0.5-1 second, wake up, read temp.

I think I'll remove the temp sensor, just send the voltage, and see if it is the temp sensing that is draining the battery.

Or I could just lick my finger and point it in the air. Or is that for wind direction?